JP2011205714A - Image processing apparatus and control method of the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology which can generate a gradation correction parameter for improving a gradation characteristic in a gradation range around a gradation value even when the frequency of the gradation value with which the gradation characteristic should be improved is lower than the frequency of other gradation values.SOLUTION: An image processing apparatus includes: a histogram generation means for generating a histogram from an input image; a detection means for detecting a gradation of interest from the histogram; a gradation correction parameter generation means for generating a gradation correction parameter for improving the gradation characteristic in the gradation range around the gradation of interest; and a correction means for correcting the gradation of the image using the gradation correction parameter. The detection means detects, as the gradation of interest, the gradation value of which the frequency is equal to or more than a predetermined threshold and the maximum value, and which has a variation of the frequency within a predetermined range including the gradation value to be smaller than a predetermined standard in the histogram.

Description

  The present invention relates to an image processing apparatus and a control method thereof.

  Conventionally, there is an image processing method for executing gradation conversion processing (for example, γ conversion processing) in accordance with the characteristics of an input image signal (hereinafter, such γ conversion processing is referred to as dynamic γ processing). Such a method is disclosed in Patent Documents 1 to 3, for example.

  Specifically, Patent Document 1 discloses a method of creating a cumulative histogram from a luminance histogram for one frame, generating a γ curve from the cumulative histogram, and performing dynamic γ processing. In Patent Documents 2 and 3, the reference values of the dark part and the bright part are set from the peak values (frequency that is the maximum value) on the low gradation side and the high gradation side in the luminance histogram of the input image signal, and the set reference values are set. A method of performing the corresponding dynamic γ processing is disclosed. In Patent Document 4, gradation correction is performed using a frequency as an upper limit value and a lower limit value for a gradation value whose frequency is equal to or higher than a predetermined upper limit value and a gradation value where the frequency is equal to or lower than a predetermined lower limit value, respectively. A technique is disclosed.

Japanese Unexamined Patent Publication No. 03-126377 Japanese Patent Laid-Open No. 08-204963 Japanese Patent Laid-Open No. 06-350873 JP 2001-125535 A

However, in the conventional method of generating the γ curve from the cumulative histogram, the gradation property of the gradation range around the gradation value having a high frequency is improved. For this reason, even if the gradation value (the gradation value of the object in the image) for which it is desirable to improve the gradation property is used, if the frequency of the gradation value is smaller than the frequencies of the other gradation values, There is a possibility that the tonality cannot be improved so much.
Specifically, as shown in FIG. 18A, a letterbox whose main gradation value is the gradation value a, an object B whose main gradation value is the gradation value b, and the main gradation value are It is assumed that an image (input image signal) including the object C having the gradation value c is input. When the frequency of the gradation value a in the luminance histogram is much higher than the frequencies of the gradation values b and c, the accumulated frequency changes greatly around the gradation value a and is accumulated around the gradation values b and c. A cumulative histogram is created so that the frequency does not change much. When the γ curve is generated from such a cumulative histogram, the gradation property of the gradation range around the gradation value a is enhanced, and the gradation property of the gradation range around the gradation values b and c is not so enhanced.

  The same applies to the method of setting the dark and bright reference values from the low and high gradation peak values. For example, when an image includes an area of the same color (for example, a letterbox or a graphic or character such as data broadcasting), the frequency of the gradation value of that area becomes large in the luminance histogram (such gradation The value shows a large peak value). For this reason, it is not possible to set an optimal reference value, and hence it is impossible to generate an optimal γ curve, and it is not possible to improve the gradation of the gradation range around the gradation value indicating a small peak value.

The same applies to the technique of performing gradation correction by limiting the frequency to a predetermined value in the luminance histogram. For example, assume that an image as shown in FIG. As shown in FIG. 18B, in the luminance histogram obtained from such an image, when the frequency of the letterbox tone value a exceeds a predetermined limit value, the tone value a The frequency is corrected to a predetermined limit value. As a result, a cumulative histogram (a histogram created by accumulating frequencies from the low gradation side to the high gradation side of the luminance histogram) as shown in FIG. 18C is obtained. The frequency of the gradation value c is relatively higher than the gradation value b. Therefore, by setting the frequency of the gradation value a as a limit value, it is possible to generate a γ curve that enhances the gradation of the gradation range around the gradation value a and the gradation range around the gradation value c. Become. However, even when such a method is used, a γ curve is generated that enhances the gradation of the gradation range around the gradation value having a relatively high frequency, and therefore the gradation around the gradation value having a relatively low frequency is generated. The gradation of the gradation range cannot be improved so much. For example, in the example of FIG. 18, the gradation of the gradation range around the gradation value b is not so improved.

  In view of the above, the present invention provides a gradation property in the gradation range around the gradation value even when the frequency of the gradation value for which the gradation property is desirable is smaller than the frequency of the other gradation values. It is an object of the present invention to provide a technique capable of generating a tone correction parameter that enhances image quality.

  An image processing apparatus according to the present invention includes a histogram creation unit that creates a histogram from an input image, and a gradation value that has a frequency that is greater than or equal to a predetermined threshold and a maximum value in the histogram, and includes a predetermined value that includes the gradation value. Detection means for detecting a gradation value whose amount of variation in frequency within a range is smaller than a predetermined reference as an attention gradation, and input / output gradation conversion for improving gradation characteristics of a predetermined gradation range including the attention gradation A gradation correction parameter generation unit that generates a gradation correction parameter having characteristics; and a correction unit that corrects the gradation of an image using the gradation correction parameter. The detection unit has a predetermined frequency in the histogram. There are multiple tone values that satisfy the condition of a tone value that is equal to or greater than the threshold value and that has a maximum value and that has a frequency variation within a predetermined range including the tone value that is smaller than a predetermined reference value. When to , And at least most gradation value of low gradation side and the most high tone gradation value of the target gradation of satisfying plurality of tone values.

  The control method of the image processing apparatus according to the present invention includes a histogram creation step of creating a histogram from an input image, and a gradation value in which the frequency is equal to or greater than a predetermined threshold and has a maximum value, Detecting step for detecting a gradation value whose amount of variation in frequency within a predetermined range including an attention level is smaller than a predetermined reference as an attention gradation, and input / output for improving the gradation of a predetermined gradation range including the attention gradation A gradation correction parameter generating step for generating a gradation correction parameter having gradation conversion characteristics; and a correction step for correcting the gradation of the image using the gradation correction parameter. In the detection step, A condition in which the frequency is a gradation value that is equal to or greater than a predetermined threshold and has a maximum value, and the variation amount of the frequency within a predetermined range including the gradation value is smaller than a predetermined reference value When the tone value satisfying there are multiple tone value of at least most gradation value of low gradation side and the most high gradation side is a target gradation of satisfying plurality of tone values.

  According to the present invention, even when the frequency of the gradation value for which it is desirable to improve the gradation is smaller than the frequency of the other gradation values, the gradation of the gradation range around the gradation value It is possible to provide a technique capable of generating a tone correction parameter that enhances the image quality.

1 is a block diagram showing a functional configuration of an image processing apparatus according to Embodiment 1. FIG. The figure which shows an example of the area | region divided | segmented into the input image. The flowchart which shows an example of the flow of a process at the time of detecting a dark part attention gradation. The figure which shows an example of a brightness | luminance histogram. The table | surface which shows each gradation value and frequency in the brightness | luminance histogram of FIG. The flowchart which shows an example of the flow of a process at the time of detecting a bright part attention gradation. The figure which shows an example of the dark part attention gradation of each area | region, and a bright part attention gradation. FIG. 5 is a diagram illustrating an example of a γ curve generated in the first embodiment. FIG. 4 is a block diagram illustrating a functional configuration of an image processing apparatus according to a second embodiment. FIG. 10 is a diagram for explaining a luminance value correction method according to the second embodiment. FIG. 9 is a block diagram illustrating a functional configuration of an image processing apparatus according to a third embodiment. The figure which shows an example of the brightness | luminance histogram of the image from which attention gradation is not detected. 10 is a flowchart illustrating an example of dark portion γ curve selection processing according to the third embodiment. The figure which shows an example of the gradation correction parameter memorize | stored beforehand. The figure which shows an example of the brightness | luminance histogram of an image with much frequency of gradation value 0. FIG. FIG. 10 is a block diagram illustrating a functional configuration of an image processing apparatus according to a fourth embodiment. The figure which shows an example of the histogram before and behind the process by a histogram pre-processing part. The figure explaining a prior art.

<Example 1>
Hereinafter, a specific example 1 of the image processing apparatus and the control method thereof according to the embodiment of the present invention will be described.
FIG. 1 is a block diagram illustrating a functional configuration of the image processing apparatus according to the present embodiment. As illustrated in FIG. 1, the image processing apparatus 100 according to the present embodiment includes a histogram creation unit 101, a target tone detection unit 102, a γ curve generation unit 103, a γ conversion unit 104, an input terminal 105, and an output terminal 106. .

  The histogram creation unit 101 creates a luminance histogram from the input image (input image signal) (histogram creation means). In this embodiment, the histogram creation unit 101 divides an input image into a plurality of regions, and creates a luminance histogram for each region. Specifically, a 1920 × 1080 image is input, and the image is divided into 4 × 4 regions (16 regions having the same size) as shown in FIG. Further, as shown in FIG. 2, the histogram creation unit 101 can distinguish each region by identification information (number or the like; blk1 to blk16 in the figure).

  In the present embodiment, it is assumed that the gradation is classified into 256 categories of gradation values 0 to 255 in the created luminance histogram. However, the number of categories is not limited to this, and may be 128 or 512, for example. Moreover, although the example of FIG. 2 shows the case where the input image is divided into 16 areas, the number of areas is not limited to this. For example, the input image may be divided into 10 × 5 and 5 × 3 regions, or may not be divided.

The target gradation detection unit 102 detects the target gradation from the luminance histogram created by the histogram creation unit 101 (detection means). Specifically, the target gradation is detected from the luminance histogram for each region. The target gradation is a gradation value for which it is desirable to improve gradation, and is a gradation value that satisfies the following conditions in the luminance histogram.
Condition 1. The frequency is greater than or equal to a predetermined threshold.
Condition 2. The frequency is a maximum value (peak value).
Condition 3. The frequency variation amount (local frequency variation amount) within a predetermined range including the gradation value (the gradation value to be determined whether or not it is the target gradation) is smaller than a predetermined reference.

Condition 1 is a condition for excluding a gradation value having an extremely low frequency as noise.
In a gradation value where it is desirable to improve gradation, the frequency appears as a peak in the luminance histogram. Condition 2 is a condition for detecting such a gradation value. Whether or not the frequency is a maximum value is compared, for example, by comparing the frequency of the target gradation value whether or not it is the target gradation and the frequency of m gradation values (m is an integer of 1 or more) before and after that. It is sufficient to judge by doing so. Note that m tone values before and after the target tone value are m tone values smaller than the target tone value and m levels larger than the target tone value. Means the key value.
In addition, when an image includes an area of the same color (same gradation value) (for example, a letterbox or a graphic or character such as data broadcast), the frequency of the gradation value of such area is the gradation. It becomes extremely larger than the frequency of the gradation values before and after the value. That is, the fluctuation amount of the frequency within a predetermined range including the gradation value becomes large (the change in the frequency from the previous gradation value and the subsequent gradation value becomes steep). Such a region (the gradation value of such a region) needs to be excluded because it is not necessary to improve the gradation. Condition 3 is a condition for excluding such an area (that is, the predetermined standard is a standard for excluding an area of the same color). Whether or not the condition 3 is satisfied depends on, for example, the frequency of the gradation value to be determined as to whether or not it is the target gradation and n gradation values before and after the gradation value (n is 1). It may be determined by comparing a value obtained by multiplying the sum of the frequencies of the above integers) by a predetermined value.

For example, i is the gradation value to be determined as to whether or not it is the target gradation, h (i) is the frequency of the gradation value i, th is the predetermined threshold, m is ra1, n is ra2, and the predetermined value is k. Then, the conditions 1 to 3 can be expressed as the following formulas 1 to 3, respectively.

  The threshold th is, for example, a value of 1% of the total frequency of one luminance histogram (that is, the total number of pixels in one area). In this embodiment, since the 1920 × 1080 image is divided into 16 regions having the same size, the threshold th is set to 1920 × 1080 ÷ 16 × 0.01 = 1296. Note that the threshold th is not limited to 1% of the total frequency, and may be any value as long as it is possible to detect a gradation value for which it is desirable to improve gradation (for example, 3% of the total frequency). And 5%).

  In Expression 2, ra1 is set to 2 in this embodiment. That is, the frequency h (i) is compared with the frequencies h (i−2), h (i−1), h (i + 1), and h (i + 2) of the two gradation values before and after the frequency h (i). It is assumed that the gradation value i satisfies the condition 2 when i) is the largest (the frequency h (i) is regarded as the maximum value). Note that ra1 is not limited to 2, but may be any value as long as it can be determined whether the frequency h (i) is a maximum value (for example, 3 or 5).

In Equation 3, in this embodiment, ra2 is 2 and k is 1.2. That is, the frequency h (i) is 1.2 times the sum of the frequencies of the two gradation values before and after the gradation value i is excluded (= 1.2 × (h (i−2) + h (i−1)). + H (i + 1) + h (i + 2))), the gradation value i satisfies the condition 3 (the fluctuation amount of the frequency within the predetermined range including the gradation value i is smaller than the predetermined reference) I reckon). Note that ra2 and k are not limited to these values, and may be any values as long as gradations of the same color region can be excluded (for example, ra2 = 3 or 5, k = 1.5 or 2). In this embodiment, the frequency h (i) is compared with the value obtained by multiplying the sum of the frequencies of n gradation values before and after the gradation value i by a predetermined value k. h (i
) And a value obtained by multiplying the sum of the frequencies of n gradation values before and after the gradation value i by a predetermined value k. Any method may be used as long as the gradation value of the region of the same color can be excluded.

  In the present embodiment, the attention gradation detection unit 102 divides the gradation into a dark part and a light part, and the dark part attention gradation (dark part attention gradation) and the bright part attention gradation (bright part attention gradation). ) Shall be detected. Specifically, gradations 2 to 126 are dark portions. A gradation value satisfying conditions 1 to 3 is searched from the low gradation side of the dark part, and a gradation value satisfying conditions 1 to 3 detected first is set as a dark part attention gradation. That is, when there are a plurality of gradation values satisfying the conditions 1 to 3, the lowest gradation value is set as the dark part attention gradation. Further, when there are no gradation values satisfying the conditions 1 to 3 in the gradations 2 to 126, the attention gradation detecting unit 102 sets the gradation 128 as the dark part attention gradation. In the present embodiment, the gradation values 0 and 1 are excluded from the search range because the frequencies of the two gradation values before and after are used for detecting the attention gradation.

  The gradation values 130 to 253 are set as bright portions. A gradation value satisfying conditions 1 to 3 is searched from the high gradation side of the bright part, and a gradation value satisfying conditions 1 to 3 detected first is set as a bright part attention gradation. That is, when there are a plurality of gradation values satisfying the conditions 1 to 3, the gradation value on the highest gradation side is set as the bright part attention gradation. In addition, when there are no gradation values satisfying the conditions 1 to 3 among the gradation values 130 to 253, the attention gradation detection unit 102 sets the gradation value 128 as the bright part attention gradation. In the present embodiment, the gradation values 254 and 255 are excluded from the search range because the frequencies of the two gradation values before and after are used for detecting the attention gradation.

  Hereinafter, the flow of processing when detecting a dark part attention gradation will be described with reference to FIGS. FIG. 3 is a flowchart showing an example of the flow of processing when detecting dark part attention gradations. FIG. 4 is a diagram illustrating an example of a luminance histogram. Hereinafter, a case where a target gradation is detected from the luminance histogram of FIG. 4 will be described. FIG. 5 is a table showing each gradation value and frequency in the luminance histogram of FIG. As described above, in this embodiment, when the dark part attention gradation is detected, the gradation value i to be processed is sequentially set from the low gradation side (gradation value 2). Note that the gradation values 2 to 13 omitted in FIG. 5 do not satisfy the conditions 1 to 3, and the processing for the gradation value 14 and later will be described in detail below.

  First, in step S11, the target gradation detection unit 102 determines whether the gradation value i is the gradation value 127 or not. When the gradation value i is the gradation value 127 (step S11: YES), the attention gradation detection unit 102 sets the gradation value 128 as the attention gradation and ends the processing. Since the gradation value 14 is not the gradation value 127 (step S11: NO), the process proceeds to step S12. In step S12, the target gradation detection unit 102 determines whether the gradation value i satisfies the condition 1 (Equation 1). The frequency of the gradation value 14 is 500 (<1296). Since the gradation value 14 does not satisfy the condition 1 (step S12: NO), the process returns to step S11. Further, when returning to step S11, the target gradation detecting unit 102 adds 1 to i. That is, after the gradation value 14, the gradation value 15 is the gradation value to be processed.

  In step S11, since it is determined that the gradation value 15 is not the gradation value 127, the process proceeds to step S12. In step S12, the frequency of the gradation value 15 is 2500, and the gradation value 15 satisfies the condition 1 (step S12: YES), so the process proceeds to step S13. In step S13, the target gradation detection unit 102 determines whether or not the gradation value i satisfies the condition 2 (Equation 2). Since the frequency 2500 of the gradation value 15 is smaller than the frequency 12000 of the gradation value 16 and the gradation value 15 does not satisfy the condition 2 (step S13: NO), the target gradation detection unit 102 adds 1 to i. Return to step S11.

In step S11, since it is determined that the gradation value 16 is not the gradation value 127, the process proceeds to step S12. Since the gradation value 16 satisfies the conditions 1 and 2 (steps S12 and 13: YES), the process proceeds to step S14. In step S14, the target gradation detection unit 102 determines whether or not the gradation value i satisfies the condition 3 (Equation 3). The sum of the frequencies of the gradation values 14, 15, 17, and 18 is 6000, and a value obtained by multiplying this value by 1.2 is 7200. Since the frequency of the gradation value 16 is greater than 7200 and the gradation value 16 does not satisfy the condition 3 (step S14: NO), the attention gradation detection unit 102 adds 1 to i, and the process returns to step S11.

  In this way, the gradation that satisfies the conditions 1 to 3 (the frequency satisfies the expressions 1 to 3) is searched. In the case of the luminance histograms shown in FIGS. 4 and 5, the gradation value 32 is first detected as the gradation satisfying the conditions 1 to 3. Therefore, the gradation value 32 is set as the dark part attention gradation.

  Next, the flow of processing when detecting a bright part attention gradation will be described with reference to FIG. FIG. 6 is a flowchart illustrating an example of a processing flow when detecting a bright part attention gradation. As described above, in this embodiment, when the bright part attention gradation is detected, the gradation value i to be processed is sequentially set from the high gradation side (gradation value 253).

  First, the target gradation detection unit 102 determines whether or not the gradation value i is the gradation value 129 (step S21). When the gradation value i is the gradation value 129 (step S21: YES), the attention gradation detection unit 102 sets the gradation value 128 as the attention gradation and ends the processing. If the gradation value i to be processed is not the gradation 129, the process proceeds to step S22. The processes in steps S22 to S24 are the same as the processes in steps S12 to S14 in FIG. When the gradation value i does not satisfy at least one of the conditions 1 to 3, the attention gradation detection unit 102 subtracts 1 from i, and the process returns to step S21.

By the method described above, the dark part attention gradation and the bright part attention gradation of each region are detected.
In this embodiment, it is assumed that the dark part attention gradation of each region is detected as shown in FIG. 7A and the bright part attention gradation is detected as shown in FIG. 7B.

  The γ curve generation unit 103 generates a gradation correction parameter (γ curve) for enhancing the gradation of the gradation range around the target gradation (gradation correction parameter generation means). In this embodiment, one γ curve is generated for the input image. Specifically, among a plurality of dark part attention gradations detected for each area, a gradation range around the lowest gradation side gradation and a plurality of bright part attention gradations detected for each area. Of these, a γ curve is generated for enhancing the gradation of the gradation range around the gradation on the highest gradation side. Thereby, it is possible to suppress blackout and overexposure.

  In this embodiment, as shown in FIG. 7A, the dark part attention gradation on the lowest gradation side (the minimum value of the dark part attention gradation) is the gradation value 25 of the region blk6. Further, as shown in FIG. 7B, the bright part attention gradation (the maximum value of the bright part attention gradation) on the highest gradation side is the gradation value 230 of the region blk2. Therefore, as shown in FIG. 8, the gradation property of the gradation range around the gradation value 25 is increased for the dark portion, and the gradation property of the gradation range around the gradation value 230 for the bright portion. A γ curve is generated to increase the value (solid line in FIG. 8). The broken line in FIG. 8 is a γ curve (conversion straight line) when the input image is output as it is. Here, an example in which the γ curve near the intermediate gradation value 127 is a linear γ curve is shown. However, the γ curve in the vicinity of the intermediate gradation value 127 may not be a linear γ curve, but may be a γ curve corresponding to the frequency in the vicinity of the intermediate gradation value 127. For example, as shown in FIG. 4, when the frequency near the intermediate gradation value 127 is large (when the frequency is larger than a predetermined threshold and has a peak value), the gradation around the intermediate gradation value 127 is obtained. It may be a γ curve that enhances the gradation of the range.

The γ conversion unit 104 corrects the gradation of the input image using the gradation correction parameter (
Correction means). Then, the corrected signal (image signal) is output to a display device (not shown). Examples of the display device include a display including a plurality of electron-emitting devices, a plasma display, a liquid crystal display, and an organic EL display. In FIG. 1, the configuration in which the image processing apparatus 100 and the display apparatus are separated is illustrated, but the image processing apparatus 100 and the display apparatus may be integrated.

  As described above, according to the present embodiment, in the case where the frequency of the gradation value (gradation value satisfying the conditions 1 to 3) for which it is desirable to improve the gradation is less than the frequency of the other gradations. Even if it exists, the gradation correction parameter for improving the gradation property of the gradation range around the gradation value can be generated. Further, the gradation value of the same color region does not satisfy the condition 3, and is excluded from the attention gradation. Thereby, an input image can be displayed with a good contrast.

  Further, in this embodiment, by generating a gradation correction parameter that enhances the gradation of the gradation range around the dark part attention gradation and the gradation range around the bright part attention gradation, the subtlety of the dark part and the bright part is generated. The contrast (contrast such as shadow) can be expressed well. Further, in the present embodiment, the input image is divided into a plurality of regions and the target gradation is detected for each region. Therefore, even if the region where the contrast needs to be expressed is a part of the image, The contrast of the portion can be expressed well.

In the present embodiment, among the plurality of target gradations detected for each region, the gradation range around the attention gradation on the lowest gradation side and the gradation range around the attention gradation on the highest gradation side. A gradation correction parameter for enhancing gradation is generated. However, the gradation correction parameter generation method is not limited to this. Blackening can be suppressed by generating gradation correction parameters to improve the gradation of the gradation range around the attention gradation on the lowest gradation side among multiple attention gradations detected for each region. can do. If a gradation correction parameter is generated to improve the gradation of the gradation range around the attention gradation on the highest gradation side among a plurality of attention gradations detected for each region, overexposure is suppressed. be able to. When the input image is not divided into a plurality of regions, in the luminance histogram of the input image, among the plurality of gradation values satisfying the conditions 1 to 3, at least the gradation value on the lowest gradation side or the most The gradation value on the high gradation side may be set as the attention gradation. Thereby, it is possible to suppress blackout and overexposure. Further, not only the gradation value on the lowest gradation side and the gradation value on the highest gradation side, but if the gradation value satisfying conditions 1 to 3 is set as the attention gradation, the conventional gradation can be improved. It is possible to improve the gradation in the gradation range around the gradation value that has not been present (a gradation value for which it is desirable to improve the gradation).
In the present embodiment, the case where the histogram creation unit 101 generates a luminance histogram has been described, but the present invention is not limited to this. For example, a histogram in which the horizontal axis is not a gradation of luminance values, but a gradation of an average value of R (red) pixel values, G (green) pixel values, and B (blue) pixel values, or an R (red) pixel value Alternatively, it may be a histogram having the maximum gradation of the G (green) pixel value and the B (blue) pixel value. Also, the gradation of one of the pixel values of the R (red) pixel value, the G (green) pixel value, and the B (blue) pixel value, such as a histogram with the horizontal axis representing the gradation of the G (green) pixel value A histogram may be used. The present embodiment can be applied to any histogram representing such image characteristics.

<Example 2>
A specific example 2 of the image processing apparatus and the control method thereof according to the embodiment of the present invention will be described below.
FIG. 9 is a block diagram illustrating a functional configuration of the image processing apparatus according to the present embodiment. As illustrated in FIG. 9, the image processing apparatus 200 according to the present embodiment includes a histogram creation unit 101, a target gradation detection unit 102, a region γ curve generation unit 201, an interpolation unit 202, an input terminal 105, and an output terminal 106. . In addition, the same code | symbol is attached | subjected to the function similar to Example 1, and the description is abbreviate | omitted.

  The area γ curve generation unit 201 generates, for each area, a gradation correction parameter for enhancing the gradation of the gradation range around the target gradation in the area (gradation correction parameter generation unit). For example, as shown in FIGS. 7A and 7B, in the region blk1, the gradation value 100 is the dark part attention gradation, and the gradation 128 is the bright part attention gradation. Therefore, for the region blk1, a γ curve is generated for enhancing the gradation of the gradation range around the gradation value 100 and the gradation range around the gradation value 128. Similarly, γ curves are generated for each of the regions blk2 to blk16 (that is, a total of 16 γ curves are generated).

  The interpolation unit 202 combines the first luminance value and the second luminance value for each pixel of the input image, and uses the combined luminance value as the final luminance value of the pixel, thereby The gradation of the image is corrected. The first luminance value is a luminance value calculated using the gradation correction parameter generated for the region (first region) to which the pixel to be processed belongs. The second luminance value is a luminance value calculated using a gradation correction parameter generated for a region (second region) adjacent to the first region. Then, the interpolation unit 202 combines the first luminance value and the second luminance value with weights based on the positional relationship between the pixel, the first region, and the second region. If the luminance value of the pixel to be processed is corrected with the tone correction parameter generated for the region to which the pixel belongs, that is, if the luminance value is corrected with a γ curve that differs for each region, the luminance between the two adjacent regions There may be a boundary line due to discontinuity of values. In this embodiment, the generation of such a boundary line can be suppressed by correcting the luminance value by the method described above.

  The second area may be one or plural. In this embodiment, all the areas adjacent to the first area are set as the second area. That is, when the region blk1 is the first region, the regions blk2, blk5, and blk6 are set as the second region. When the region blk6 is the first region, the regions blk1, blk2, blk3, blk5, blk7, blk9, blk10, and blk11 are the second regions.

なお、上述した合成率はテーブルや数式から得られるものとする。 Hereinafter, a specific example will be described with reference to FIG.
As shown in FIG. 10, the luminance value of the pixel P belonging to the region blk1 is corrected by using four γ curves respectively corresponding to the region blk1 and the surrounding regions blk2, blk5, blk6. Specifically, using the horizontal distance x and the vertical distance y from the center coordinate of the region blk1 to the position of the pixel P, the luminance values obtained from these γ curves are weighted and synthesized. The composite rate when the distance x = x1 is 0.7, and the composite rate when the distance y = y1 is 0.6. For the regions blk1, blk2, blk5 and blk6, the γ curves generated using the distances from the center coordinates of each region to the position of the pixel P are denoted by γ1, γ2, γ5 and γ6, respectively. When the luminance values after correction when the luminance value of the pixel P is corrected by each γ curve are OP1, OP2, OP5, and OP6, the final luminance value OP of the pixel P is expressed by Expression 4. .

Note that the above-described synthesis rate is obtained from a table or a mathematical expression.

  As described above, according to the present embodiment, tone correction parameters are generated for each region, so that tone correction (dynamic γ processing) according to the feature of each region can be performed. In other words, it is possible to express a good contrast by enhancing the gradation of the dark portion for the region having the attention gradation in the dark portion and sinking the dark portion for the region having no attention gradation in the dark portion. It is possible to express a good contrast by enhancing the gradation of the bright part for the area with the attention gradation in the bright part and floating the bright part for the area without the attention gradation in the bright part. . In this embodiment, the first luminance value and the second luminance value described above are combined to obtain a final luminance value, that is, a luminance value in consideration of the gradation correction parameter of the adjacent region is obtained. Thus, it is possible to suppress the generation of a boundary line between two adjacent regions.

  In this embodiment, all the areas adjacent to the first area are set as the second area. However, the method for selecting the second area is not limited to this. Four regions adjacent to the first region in the vertical and horizontal directions may be used as the second region. The effect mentioned above can be acquired also by it. Also, one region adjacent to the first region in any of the upper, lower, left and right directions may be set as the second region. If at least one region adjacent to any one of the upper, lower, left, and right sides of the first region is set as the second region, the generation of a boundary line between the two regions adjacent in one direction can be suppressed.

Moreover, the occurrence of a boundary line can be suppressed by considering the tone correction parameter in the region close to the pixel to be processed among the regions adjacent to the first region. For example, the second region may be selected according to the position of the pixel to be processed in the first region. Specifically, when the first region is divided into four 2 × 2 regions, and the pixel to be processed belongs to the lower right region of the four regions, the right of the first region, The area adjacent to the lower and lower right may be the second area.
Further, although the case of using a luminance histogram has been illustrated, the present embodiment can be applied to any histogram that represents image characteristics, such as a histogram in which the horizontal axis is a gradation of pixel values other than the luminance value. When a pixel value other than the luminance value is used, the interpolation unit 202 corrects the pixel value instead of the luminance value. The pixel values are R (red) pixel values, G (green) pixel values, B (blue) pixel values, R (red) pixel values, G (green) pixel values, and B (blue) pixel values. Values representing the color and luminance of the pixels, such as the average value and luminance value of the pixel.

<Example 3>
A specific example 3 of the image processing apparatus and its control method according to the embodiment of the present invention will be described below. Specifically, even when an image in which a gradation value (attention gradation) satisfying the conditions 1 to 3 described in the first embodiment cannot be detected, a gradation correction parameter suitable for the image is set. How to do will be described.

  First, an image in which the target gradation cannot be detected will be described. Such an image is, for example, an image composed of a combination of regions of the same color as an animation. In such an image, a luminance histogram as shown in FIG. 12 is obtained. In such a luminance histogram, since the image is composed of a combination of regions of the same color, the frequency variation from the preceding and following gradation values to the peak gradation value becomes severe (arrows in FIG. 12). In other words, in such an image, there is no gradation value satisfying the condition 3, so that the attention gradation is not detected. As a result, even in the case of a dark image, the gradation of the dark part cannot be improved and blackout occurs. Further, even in a bright image, the gradation of the bright part cannot be improved, and whiteout occurs.

  FIG. 11 is a block diagram illustrating a functional configuration of the image processing apparatus according to the present embodiment. As illustrated in FIG. 11, the image processing apparatus 300 according to the present embodiment includes a histogram creation unit 101, a target gradation detection unit 102, a frequency calculation unit 301, a region γ curve selection unit 302, and an interpolation unit 202. In addition, the same code | symbol is attached | subjected to the function similar to Example 1 and Example 2, and the description is abbreviate | omitted.

  The frequency calculation unit 301 calculates the sum of the frequencies from the minimum gradation value of the luminance histogram to a predetermined gradation value as the frequency (dark part frequency) on the low gradation side (dark part) (calculation unit). Further, the sum of the frequencies from the predetermined gradation value to the maximum gradation value of the luminance histogram is calculated as the frequency (bright portion frequency) on the high gradation side (bright portion). In this embodiment, as shown in FIG. 12, the integrated value a of the frequency from the gradation value 0 to the gradation value 127 is set as the dark portion frequency, and the integrated value b of the frequency from the gradation value 128 to the gradation value 255 is set. Brightness frequency.

The region γ curve selection unit 302 has a function of selecting any one of a plurality of gradation correction parameters stored in a storage unit (storage unit) (not shown) in addition to the function of the region γ curve generation unit 201 of the second embodiment. Have Specifically, the area γ curve selection unit 302 has a plurality of gradation correction parameters stored according to the dark part frequency and the bright part frequency when the target gradation detection unit 102 does not detect the target gradation. Select one of the following. In the present embodiment, it is assumed that the tone correction parameter is determined for each region as in the second embodiment.
The interpolation unit 202 corrects the tone of the input image using the selected tone correction parameter when the target tone detection unit 102 does not detect the target tone. Note that the gradation correction method in this embodiment is the same as that in the second embodiment.

  The plurality of gradation correction parameters stored in the storage unit are, for example, a plurality of gradation correction parameters corresponding to the dark part frequency and the light part frequency. In this embodiment, as shown in FIG. 14, dark portion γ curves B (0) to B (4) with gradation values 0 to 128, and bright portion γ curves W (with gradation values 192 to 255). It is assumed that 0) to W (4) are stored. In the example of FIG. 14, dark portion γ curves B (4), B (3), B (2), B (1), and B (0) are set in order of increasing gradation on the low gradation side. Yes. In addition, bright portion γ curves W (4), W (3), W (2), W (1), and W (0) are set in order of increasing gradation on the high gradation side. The region γ curve selection unit 302 selects a dark part γ curve and a bright part γ curve, respectively. The γ curve from the gradation value 128 to the gradation value 191 may be generated by connecting the dark portion γ curve and the bright portion γ curve with a straight line, or a function that does not become discontinuous at the gradation values 128 and 191. May be generated by connecting the dark portion γ curve and the bright portion γ curve. Thereby, one γ curve is finally generated.

  Hereinafter, the process performed by the region γ curve selection unit 302 will be described in detail. In particular, a process for selecting a dark part γ curve (dark part γ curve selection process) will be described in detail. FIG. 13 is a flowchart illustrating an example of a processing flow until the region γ curve selection unit 302 selects a dark portion γ curve.

  First, the region γ curve selection unit 302 determines whether or not a target gradation is detected by the target gradation detection unit 102 (step S31). If the target gradation is detected (step S31: YES), the process proceeds to step S33. When the target gradation is not detected (step S31: NO), the process proceeds to step S32. In step S33, a tone correction parameter is generated as in the second embodiment, and the process is terminated. In step S <b> 32, the region γ curve selection unit 302 compares the dark portion frequency calculated by the frequency calculation unit 301 with the threshold Bth. That is, it is determined how dark the image of the divided area is.

  In this embodiment, the threshold value Bth is set to 1/16 of the total number of pixels in the region. In the present embodiment, 1920 × 1080 is divided into 16 as in Embodiment 1 and Embodiment 2, so the total number of pixels in one divided area is (1920 × 1080) ÷ 16 = 129600 pixels. . 1/16 of the total number of pixels, that is, 8100 is the threshold value Bth. When the dark part frequency is equal to or greater than the threshold Bth (for example, the dark part frequency is 10,000) (step S32: Yes), the region γ curve selection unit 302 determines that the image of the processing target region is a dark image. Therefore, the region γ curve selection unit 302 selects the dark part γ curve B (1) for improving the gradation property on the low gradation side, and ends the processing (step S34). When the dark area frequency is less than the threshold Bth (for example, the dark area frequency is 1000) (step S32: NO), the region γ curve selection unit 302 determines that the image of the region to be processed is not a dark image. Therefore, the region γ curve selection unit 302 selects the dark part γ curve B (0) that does not increase the gradation on the low gradation side so much, and ends the process (step S35).

The process of selecting the bright part γ curve is the same as selecting the dark part γ curve. Specifically, when the target gradation detection unit 102 detects the target gradation, a gradation correction parameter is generated as in the second embodiment. When the target gradation detection unit 102 does not detect the target gradation, the bright part frequency is compared with the threshold value. When the bright part frequency is equal to or greater than the threshold value, it is determined that the image in the processing target area is a bright image, and a bright part γ curve (for example, a bright part) for improving the gradation on the high gradation side so as not to be overexposed Part γ curve W (4)) is selected. If the bright part frequency is less than the threshold value, it is determined that the image in the processing target area is not a bright image, and the bright part γ curve (for example, the bright part γ curve W (2 )) Is selected.
Then, the interpolation unit 202 interpolates the luminance value output from the γ curve for each region as in the second embodiment.

  As described above, according to the present embodiment, even when an image in which a target gradation cannot be detected, such as an image composed of a combination of regions of the same color such as animation, is input, Since the γ curve is determined according to the above, good contrast can be expressed.

In the present embodiment, the case where the input image is divided into a plurality of regions in the same manner as in the first and second embodiments has been described. However, even if the input image is not divided, the present embodiment. The configuration can be applied. Specifically, the gradation correction parameter may be selected according to the dark part frequency and the light part frequency of the entire image.
In this embodiment, the method of the second embodiment is applied as the gradation correction method. However, the gradation may be corrected by the method of the first embodiment.

  In this embodiment, the dark portion γ curves B (1) and B (0) are selected according to the dark portion frequency, and the bright portion γ curves W (4) and W (2) are selected according to the light portion frequency. However, the γ curve to be selected is not limited to this. Any one of a plurality of dark part γ curves stored in advance according to the dark part frequency may be selected, or any one of a plurality of bright part γ curves stored in advance may be selected according to the light part frequency. . In this embodiment, five dark part γ curves and five bright part γ curves are stored, but the number is not limited to this. Three or ten dark portion γ curves and light portion γ curves may be stored. The number of dark part γ curves and bright part γ curves may not be the same. The γ curve may not be stored separately for the dark part and the bright part.

In this embodiment, two γ curves are selected in accordance with the dark part frequency and the bright part frequency, but such a configuration is not necessary. The γ curve may be selected according to at least the dark part frequency or the bright part frequency. If the γ curve is selected according to the dark area frequency, the contrast of the dark area can be expressed well. If the γ curve is selected according to the brightness of the bright part, the contrast of the dark part can be expressed well.
Further, although the case of using a luminance histogram has been illustrated, the present embodiment can be applied to any histogram that represents image characteristics, such as a histogram in which the horizontal axis is a gradation of pixel values other than the luminance value.

<Example 4>
A specific example 4 of the image processing apparatus and the control method thereof according to the embodiment of the present invention will be described below. Specifically, even when an image having an 8-bit (256) gradation and a large gradation value of 0 or gradation value 255 is input, a gradation correction parameter suitable for the image is set. Will be described.

  For example, when a night scene is shot with a digital camera, an image with a high gradation value of 0 may be obtained, and a luminance histogram as shown in FIG. 15 may be obtained. In the method described in the first and second embodiments, the gradation values 0 and 1 are excluded from the search range when searching for the target gradation in the dark portion. For this reason, in the luminance histogram having the frequency peak at the gradation value 0 as described above, the gradation value 0 (or a gradation in the vicinity thereof) is not detected as the dark part attention gradation, and blackout occurs. Further, even in a bright image (an image having a frequency peak at the gradation value 255), the gradation property of the bright portion (gradation property around the gradation value 255) cannot be improved, and whiteout occurs. End up.

  FIG. 16 is a block diagram illustrating a functional configuration of the image processing apparatus according to the present embodiment. As illustrated in FIG. 16, the image processing apparatus 400 according to the present embodiment includes a histogram creation unit 101, a histogram preprocessing unit 401, a target tone detection unit 102, a γ curve generation unit 103, a γ conversion unit 104, and an input terminal 105. And an output terminal 106. In addition, the same code | symbol is attached | subjected to the function similar to Example 1, and the description is abbreviate | omitted.

  The histogram preprocessing unit 401 analyzes the luminance histogram created by the histogram creation unit 101, and rewrites and outputs the luminance histogram data when the frequency of the gradation value 0 or the gradation value 255 is large (histogram preprocessing). means). The target gradation detection unit 102 detects the target gradation from the luminance histogram data output from the histogram preprocessing unit 401. As described in the first exemplary embodiment, the target gradation detection unit 102 divides the gradation into a dark part and a bright part, and the dark part attention gradation (dark part attention gradation) and the bright part attention gradation (bright part attention). ) Is detected.

  Hereinafter, the flow of the rewriting process for the gradation of the dark part of the luminance histogram will be specifically described. FIG. 17A shows the frequency of each gradation value in the luminance histogram shown in FIG. The histogram preprocessing unit 401 determines whether or not the frequency with the gradation value 0 is larger than the frequency with the gradation value 2. As shown in FIG. 17A, when the frequency of gradation value 0 (3500) is higher than the frequency of gradation value 2 (2300), the frequency of gradation value 2 (2300) is set to the gradation value 0. Rewrite to frequency (3500). FIG. 17B is a diagram illustrating data of the luminance histogram after being rewritten by the histogram preprocessing unit 401. Note that the histogram preprocessing unit 401 does not perform the process of rewriting the frequency of the gradation value 2 when the frequency of the gradation value 0 is smaller than the frequency of the gradation value 2.

  Similarly, the histogram preprocessing unit 401 performs a rewriting process on the gradation of the dark part of the luminance histogram. Specifically, it is determined whether or not the frequency of the gradation value 255 is greater than the frequency of the gradation value 253. If the frequency of the gradation value 255 is greater than the frequency of the gradation value 253, the gradation value The frequency of 253 is rewritten to the frequency of the gradation value 255. Note that when the frequency of the gradation value 255 is not higher than the frequency of the gradation value 253, the histogram preprocessing unit 401 does not perform the process of rewriting the frequency of the gradation value 253.

  The target gradation detection unit 102 performs processing for detecting the target gradation from the luminance histogram rewritten by the histogram preprocessing unit 401. Therefore, even when an image having a frequency peak at the gradation value 0 or the gradation value 255 is input, the attention gradation detection unit 102 converts the gradation value 2 or gradation value 253 to the attention gradation. Can be detected as.

As described above, according to the present embodiment, when an image having an 8-bit gradation and a large number of gradation values 0 (minimum gradation value) and gradation value 255 (maximum gradation value) is input. Even so, it is possible to determine an appropriate γ curve (a γ curve having the gradation value 2 or the gradation value 253 as the target gradation). For this reason, it is possible to apply a tone correction parameter suitable for an image, and to realize a good contrast. Note that the case where the input image has an 8-bit gradation has been described as an example, but the present invention can be applied regardless of how many bits the input image has. Here, the case where the target gradation detection unit 102 compares the frequency of the target gradation value with the frequencies of the two gradation values before and after removing the gradation value has been described. It is not limited. That is, the present invention can be applied to the case where the frequency of the target gradation value is compared with the frequency of n gradation values (n is an integer of 1 or more) before and after the gradation value is excluded. If the frequency of the minimum gradation value (gradation value 0) is greater than the frequency of the gradation value of the minimum gradation value + n, the histogram preprocessing unit 401 minimizes the frequency of the gradation value of the minimum gradation value + n. What is necessary is just to rewrite to the frequency of a gradation value. When the frequency of the maximum gradation value (gradation value 255 in the case of 8-bit gradation) is greater than the frequency of the gradation value of maximum gradation value-n, the gradation value of maximum gradation value-n May be rewritten to the frequency of the maximum gradation value. In this embodiment, the method of the first embodiment is applied as the gradation correction method. However, the gradation may be corrected by the method of the second embodiment. In addition, even when the input image is divided into a plurality of regions and a luminance histogram for each region is created, the input image is not divided into a plurality of regions and the entire image is created with a luminance histogram. Even if it exists, a present Example is applicable. When the image is not divided, the gradation correction parameter may be determined according to the luminance histogram of the entire image.
Further, although the case of using a luminance histogram has been illustrated, the present embodiment can be applied to any histogram that represents image characteristics, such as a histogram in which the horizontal axis is a gradation of pixel values other than the luminance value.
The histogram pre-processing unit 401 may be configured to set whether or not to use a function of rewriting luminance histogram data. When the setting is made so that the function of the histogram preprocessing unit 401 is not used, as described in the first embodiment, the histogram generated by the histogram creation unit 101 is input to the target gradation detection unit 102 as it is.

As described above, according to the image processing apparatus and the control method thereof according to the first to fourth embodiments, it is possible to detect a gradation value for which it is desirable to improve gradation as the attention gradation. Thereby, even when the frequency of the target gradation is smaller than the frequency of other gradation values, it is possible to generate a gradation correction parameter that improves the gradation of the gradation range around the target gradation. .
In addition, the said Examples 1-4 can be combined.

DESCRIPTION OF SYMBOLS 100 Image processing apparatus 101 Histogram creation part 102 Attention gradation detection part 103 γ curve generation part 104 γ conversion part

Claims (18)

A histogram creating means for creating a histogram from the input image;
In the histogram, a gradation value whose frequency is equal to or greater than a predetermined threshold and has a maximum value, and whose variation amount of the frequency within a predetermined range including the gradation value is smaller than a predetermined reference is a target gradation value. Detecting means for detecting as,
A gradation correction parameter generating means for generating a gradation correction parameter having an input / output gradation conversion characteristic that enhances the gradation of a predetermined gradation range including the target gradation;
Correction means for correcting the gradation of the image using the gradation correction parameter;
Have
In the histogram, the detection means has a gradation value whose frequency is equal to or greater than a predetermined threshold and has a maximum value, and a variation amount of the frequency within a predetermined range including the gradation value is smaller than a predetermined reference. When there are a plurality of gradation values that satisfy the condition, the at least the lowest gradation value and the highest gradation value among the plurality of gradation values satisfying the condition are An image processing apparatus characterized by having a key. The detection means is a frequency of n gradation values (n is an integer greater than or equal to 1) before and after the gradation value that is the target of determination as to whether or not to select the target gradation. 2. The image processing apparatus according to claim 1, wherein the amount of variation in the frequency within the predetermined range is considered to be smaller than a predetermined reference when the sum is less than a value obtained by multiplying the total by a predetermined value. When the histogram created by the histogram creating means is analyzed and the frequency of the minimum gradation value is larger than the frequency of the gradation value of the minimum gradation value + n, the frequency of the gradation value of the minimum gradation value + n is minimized. When the frequency of the maximum gradation value is larger than the frequency of the gradation value of the maximum gradation value-n, the frequency of the gradation value of the maximum gradation value-n is changed to the maximum gradation value. The image processing apparatus according to claim 2, further comprising a histogram preprocessing unit that rewrites the frequency to The gradation correction parameter generation means generates, as the gradation correction parameter, a γ curve having an input / output gradation conversion characteristic that enhances gradation of a predetermined gradation range including the target gradation. The image processing apparatus according to claim 1. The histogram creating means divides the image into a plurality of regions, creates a histogram for each region,
The detecting means detects a target gradation from the histogram for each region,
The gradation correction parameter generating means generates a gradation correction parameter having an input / output gradation conversion characteristic for improving the gradation of a predetermined gradation range including a target gradation of the area for each area,
For each pixel, the correction means calculates a pixel value calculated using a tone correction parameter generated for the first region to which the pixel belongs, and a second region adjacent to the first region. The pixel value calculated using the tone correction parameter generated for the pixel is synthesized with a weight based on the positional relationship between the pixel and the first region and the second region. The image processing apparatus according to claim 1, wherein a gradation of the image is corrected by setting a pixel value as a final pixel value of the pixel. The histogram creating means divides the image into a plurality of regions, creates a histogram for each region,
The detecting means detects a target gradation from the histogram for each region,
The gradation correction parameter generation unit is configured to perform gradation characteristics of a predetermined gradation range including at least the attention gradation on the lowest gradation side among the plurality of attention gradations detected for each region for the image. The image processing apparatus according to claim 1, wherein a gradation correction parameter having an input / output gradation conversion characteristic that enhances the image quality is generated. The histogram creating means divides the image into a plurality of regions, creates a histogram for each region,
The detecting means detects a target gradation from the histogram for each region,
The gradation correction parameter generation means has a gradation property of a predetermined gradation range including at least the attention gradation on the highest gradation side among the plurality of attention gradations detected for each region for the image. The image processing apparatus according to claim 1, wherein a gradation correction parameter having enhanced input / output gradation conversion characteristics is generated. Storage means for storing a plurality of predetermined gradation correction parameters;
A calculation means for calculating the sum of the frequencies from the minimum gradation value of the histogram to the predetermined gradation value as the frequency on the low gradation side;
When the target gradation is not detected by the detection means,
The gradation correction parameter generation means selects one of the plurality of gradation correction parameters according to the frequency on the low gradation side;
The image processing apparatus according to claim 1, wherein the correction unit corrects the gradation of the image using the selected gradation correction parameter. Storage means for storing a plurality of predetermined gradation correction parameters;
A calculation means for calculating the sum of the frequencies from the predetermined gradation value to the maximum gradation value of the histogram as the frequency on the high gradation side;
When the target gradation is not detected by the detection means,
The gradation correction parameter generation means selects any of the plurality of gradation correction parameters according to the frequency on the high gradation side,
The image processing apparatus according to claim 1, wherein the correction unit corrects the gradation of the image using the selected gradation correction parameter. A histogram creation step for creating a histogram from the input image;
In the histogram, a gradation value whose frequency is equal to or greater than a predetermined threshold and has a maximum value, and whose variation amount of the frequency within a predetermined range including the gradation value is smaller than a predetermined reference is a target gradation value. Detecting step to detect as,
A gradation correction parameter generation step of generating a gradation correction parameter having an input / output gradation conversion characteristic that enhances gradation of a predetermined gradation range including the target gradation;
A correction step of correcting the gradation of the image using the gradation correction parameter;
Have
In the detection step, in the histogram, a gradation value whose frequency is equal to or greater than a predetermined threshold and has a maximum value, and a variation amount of the frequency within a predetermined range including the gradation value is smaller than a predetermined reference When there are a plurality of gradation values that satisfy the condition, the lowest gradation value and the highest gradation value among the plurality of gradation values satisfying the condition are the attention floor. A method for controlling an image processing apparatus, characterized by comprising: In the detection step, the frequency of the gradation value to be determined as to whether or not to be the target gradation is a frequency of n gradation values (n is an integer of 1 or more) before and after the gradation value is excluded. The image processing apparatus according to claim 10, wherein when the sum is less than a value obtained by multiplying a sum by a predetermined value, a variation amount of the frequency within the predetermined range is regarded as being smaller than a predetermined reference. Control method. When the histogram created by the histogram creating step is analyzed and the frequency of the minimum gradation value is larger than the frequency of the gradation value of the minimum gradation value + n, the frequency of the gradation value of the minimum gradation value + n is minimized. When the frequency of the maximum gradation value is larger than the frequency of the gradation value of the maximum gradation value-n, the frequency of the gradation value of the maximum gradation value-n is changed to the maximum gradation value. The method of controlling an image processing apparatus according to claim 11, further comprising a histogram preprocessing step of rewriting to a frequency of. In the gradation correction parameter generation step, a γ curve having an input / output gradation conversion characteristic that enhances the gradation of a predetermined gradation range including the target gradation is generated as the gradation correction parameter. The method for controlling an image processing apparatus according to claim 10, wherein the image processing apparatus has a control method. In the histogram creation step, the image is divided into a plurality of regions, and a histogram for each region is created.
In the detection step, a target gradation is detected from the histogram for each region,
In the gradation correction parameter generation step, for each area, a gradation correction parameter having an input / output gradation conversion characteristic that enhances the gradation of a predetermined gradation range including a target gradation of the area is generated;
In the correction step, for each pixel, a pixel value calculated using the gradation correction parameter generated for the first region to which the pixel belongs and a second region adjacent to the first region The pixel value calculated using the tone correction parameter generated for the pixel is synthesized with a weight based on the positional relationship between the pixel and the first region and the second region, and synthesized. The method of controlling an image processing apparatus according to claim 10, wherein a gradation of the image is corrected by setting a pixel value to a final pixel value of the pixel. . In the histogram creation step, the image is divided into a plurality of regions, and a histogram for each region is created.
In the detection step, a target gradation is detected from the histogram for each region,
In the gradation correction parameter generation step, gradation characteristics of a predetermined gradation range including at least the attention gradation on the lowest gradation side among the plurality of attention gradations detected for each region in the image 14. The method of controlling an image processing apparatus according to claim 10, wherein a gradation correction parameter having an input / output gradation conversion characteristic that enhances image quality is generated. In the histogram creation step, the image is divided into a plurality of regions, and a histogram for each region is created.
In the detection step, a target gradation is detected from the histogram for each region,
In the gradation correction parameter generation step, the gradation of a predetermined gradation range including at least the attention gradation on the highest gradation side among the plurality of attention gradations detected for each region is obtained for the image. The method of controlling an image processing apparatus according to claim 10, wherein a gradation correction parameter having enhanced input / output gradation conversion characteristics is generated. A calculation step of calculating the sum of the frequencies from the minimum gradation value of the histogram to a predetermined gradation value as the frequency on the low gradation side;
When the target gradation is not detected by the detection step,
In the gradation correction parameter generation step, one of a plurality of predetermined gradation correction parameters stored in the storage unit is selected according to the frequency on the low gradation side,
The image processing apparatus control method according to claim 10, wherein in the correction step, the gradation of the image is corrected using the selected gradation correction parameter. A calculation step of calculating the sum of the frequencies from the predetermined gradation value to the maximum gradation value of the histogram as the frequency on the high gradation side;
When the target gradation is not detected by the detection step,
In the gradation correction parameter generation step, one of a plurality of predetermined gradation correction parameters stored in the storage unit is selected according to the frequency on the high gradation side,
The image processing apparatus control method according to claim 10, wherein in the correction step, the gradation of the image is corrected using the selected gradation correction parameter.

Images